The present paper is a quasi-experimental research and uses pre-test and post-test method with control group. The tool used in this paper in the pre-test and post-test stages is Cooper Smith's Self-Esteem Questionnaire. Sub-scales include educational self-esteem, family self-esteem, physical self-esteem, and social self-esteem. In this paper, group narrative therapy was held in eight weekly sessions for 90 minutes and 15 participants. The statistical universe comprises all 10-14 years old girls of Torkaman Welfare Organization in District 17 of Tehran. Multi-stage random sampling method was used in this paper. All these girls were given Cooper Smith's self-esteem test. From among the girls who had lower self-esteem, 15 girls were selected randomly as the test group and 13 ones were selected as control group. This paper has used t-student test, Levine F test for equality of variances, box test for model goodness of fit, covariance test for calculating the effect between the subjects in two control and test groups.

The main result of this paper indicates that narrative therapy has a significant effect on general self-esteem, social self-esteem, and physical self-esteem; but it has no significant effect on family self-esteem and educational self-esteem.